LaunchpadX
About Event
LaunchPadX National Innovation Hackathon 2026 is a 48-hour offline national hackathon held on August 22–23, 2026, at the Sreenidhi Institute of Science and Technology (SNIST), Hyderabad. Organized by VetriaAI, it brings together engineering students and developers to build real-world AI and tech solutions.
Event Timeline & Schedule
Check chronological milestones and event slots.
No schedule events are configured for this hackathon.
Event Problem Statements & Tracks
Explore available challenge tracks and technical problem briefs.
The Agent Hub
Build an autonomous AI agent system that can Plan, Use Tools, Remember context, and Execute multi-step tasks. System Requirements (ALL are required): • Planning & Decision Making: Must perform multi-step reasoning and task decomposition. • Tool Usage: Must interact with external APIs, the web, databases, or file systems. • Memory Management: Must implement session/user profile memory or vector DB retrieval. • Autonomous Execution: Must complete multi-step objectives with minimal human input. Strictly Prohibited: Stateless, single-prompt-to-single-response chatbots.
The Prediction Machine
Build a full machine learning pipeline that trains on structured data and produces meaningful predictions. Pipeline: Structured Dataset → Model Training → Evaluation Metrics → Prediction Output System Requirements (ALL are required): • Dataset: Must use structured data (CSV, SQL, time-series, or sensor data). • Training Engine: Must build a custom-trained model — Random Forest, XGBoost, Neural Nets, LSTM, or similar (no LLM-generated outputs). • Evaluation: Must report measurable performance — RMSE, R², Accuracy, or F1. • Output: Must produce a meaningful prediction (e.g., forecasting, anomaly/fraud detection, classification).
Computer Vision
Build a real CV pipeline that processes image or video input and produces measurable, structured output. System Requirements (ALL are required): • Input Stream: Must process image or video input. • Vision Pipeline: Must use a real CV architecture — CNN, YOLO, ResNet, or Vision Transformer (no black-box API calls as the "pipeline"). • Automated Understanding: Must perform object detection, classification, segmentation, or OCR. • Measurable Output: Must produce bounding boxes, labels, masks, or extracted text — something quantifiable, not just a text description.
Cross-Sense AI
Build a multimodal AI system that fuses at least 2 distinct input modalities and produces a unified output. System Requirements (ALL are required): • Multi-Input: Must use a minimum of 2 distinct modalities (e.g., Audio + Text, Vision + Sensor). • Cross-Modal Interaction: Modalities must actively influence each other, not run in parallel. • Contextual Reasoning: Must generate understanding across the different inputs together. • Unified Output: Must produce a single cohesive report, decision, or classification. Strictly Prohibited: Building separate, isolated AI modules that never interact with one another.
Offline-First Systems
Build a resilient, offline-first system that operates without internet connectivity and coordinates across multiple nodes. Pipeline: Real-World Input → Local/Edge Processing → Real-Time Decision → Coordinated Output System Requirements (ALL are required): • No-Connectivity Operation: Must work without internet/cellular — using only device-native capability (Bluetooth LE, Wi-Fi Direct, local compute). • Multi-Node Coordination: Must coordinate across multiple devices/feeds/paths, and handle a node dropping or a path failing mid-operation. • Real-Time Constraint: Must process and respond within a bounded, self-reported latency. • Noise & Conflict Resolution: Must handle messy real input — duplicates, occlusion, and conflicting routes — not clean, ideal data. Strictly Prohibited: Demoing only the "happy path" without showing failure/recovery behavior.
Prizes & Categories
Review available tracks and reward distributions.
Top performing team overall
Second place overall
Third place overall
Fourth place overall
Fifth place overall
FAQ & Queries
Find answers to common developer inquiries.
No FAQs configured.
Total Prize Pool
1,00,000
Timeline
August 22, 2026
to August 24, 2026
Location
SNIST
Registrations
934 Competitors
Tickets & Passes